Airplanes are like really big candy bars
Dag Bennett, Helen Aston
The Ehrenberg Centre for Research in Marketing
Our Contribution
Our new finding is that the competitive structure of a capital
goods market conforms to the laws of marketing
1. Contrary to contemporary theory (e.g. McCabe & Stern,
2009, Stacey & Wilson, 2015) we show that the market
structure is like that for consumer goods
2. Buying is stochastic, but predictable based on underlying
buying propensities
3. Readily available, and reliable information makes analysis
accessible and straightforward
Large Commercial Aircraft
Conceptual Framework
Brand performance metrics are predictable for capital goods.
Commercial
Aircraft brand
Market
Share %
Penetration
(observed) %
Penetration
Purchases
(theoretical) % per buyer
(observed)
Purchases
per buyer
(theoretical)
Boeing
46
86
85
101
97
Airbus
43
82
81
95
94
Canadair
4
18
18
47
41
Embraer
4
16
16
37
34
Other
3
16
16
31
30
20
44
43
62
59
Average Brand
1900 aircraft sold each year
$80-380 million each
- 20 year life
- multi-year purchasing process
Planes are not FMCG
Drawn out purchasing
Final prices and services are negotiated
Buying centres, Dozens of participants
We show that
.
This is the well-known double jeopardy pattern, where big
brands have more customers, who buy more often
Aircraft makers share customers
Airbus
We have great products,
but now it’s clear we
can market them better.
John Leahy, COOCustomers
NBD-Dirichlet model (Goodhardt, Ehrenberg & Chatfield,
1984) of buying behaviour.
Validated by application to many different FMCG categories,
consumer durables and services, leading to empirical
generalizations through replication and extension.
.
This illustrates the duplication of purchase law—brands
share their customers more with other big brands than with
smaller brands
Existing Knowledge
Implications for Researchers
1. Double Jeopardy Law (DJ) ---- Big brands have more
customers, who are slightly more loyal
2. Duplication of Purchase Law (DoP) ---- Brands share
customers in line with the size of the other brands
3. Portfolio loyalty is the norm
4. Brand shares are stable
5. Penetration leads to growth and raises brand loyalty
6. Brand loyalty is a function of brand size
Brand performance needs to be analysed in context, using
established knowledge
Interpret BPMs (loyalty, switching, share of buyers) in
reference to established patterns
Predictive power of stochastic models of buyer behavior
provide a rich context to study buyer/seller interactions
Market Structure
Nerd Power
The data was from Planespotters.net– a world wide network
of aviation nerds who are very keen to know exactly which
airline has which plane—each of which has a unique code
number, and a unique charm to the planespotters.
Implications for Managers
Sales increases come from attracting new customers across
the board and stochastic modeling allows managers to
predict sales success. This can use off-the-shelf software,
readily available data, and simple analysis techniques
Frequency of purchase doesn’t vary much between
suppliers, so strategies to capture more of a company’s
capital allocations won’t work
But strategies to increase market penetration will
This research makes us
question all our thinking
about industrial market
behaviour.
Prof. Øyvind Bøhren,
Industry consultant